Machine learning domain adaptation in spin models with continuous phase transitions.

The main question raised in the article is whether a neural network trained on a spin lattice model in one universality class can be used to test a model in another universality class. The quantities of interest are the critical phase transition temp…

Preview only. Read the full abstract at the source